Single unit activity of muscle spindle afferents has been recorded in the awake monkey's dorsal root ganglia and indicates that, during slow pursuit handtracking, gamma fusimotor neurons may fire while alpha skeletomotor neurons are silent. Since all the cerebellar neurons recorded in dentate and interposed nuclei and half the neurons in motor cortex fired in patterns identical to those of spindle afferents (and by inference, gamma motor neurons), and since both the cerebellum and motor cortex are known to influence gamma motor neurons, our hypothesis is that they may cause the dissociation. Since alpha-gamma dissociation has been seen in this overtrained movement, and since it has not been seen to this degree in the untrained movements of man and various animals, it seems plausible that it is enhanced by training. Two ways in which the dissociation might improve motor performance are by stiffening the wrist joint against unwanted small perturbations by increasing the activity of the segmental stretch reflex, and/or by diminishing the tendency for tremor by reducing the synchronizing effect that spindle afferents may have on motor unit discharge. Three sets of experiments are herein proposed to reveal the mechanism of the alpha-gamma dissociation and its use in motor control. One set of experiments will combine EMG, single unit recording, and spike triggered averaging with a lesion of dentate, interpositus, or peripheral nerve to test our hypothesis on the cause of alpha-gamma dissociation and its effect on motor performance. A second experiment will consist of recording the activity of Purkinje cells during the development of alpha-gamma dissociation to see if their firing patterns support the theories of a cerebellar role in adaptive motor control. A third set of experiments will consist of a search for alpha-gamma dissociation in other trained movements to test our hypothesis on its generality and behavioral use. During the slow tracking, cerebellar neurons and spindle afferents discharged in bidirectional patterns lacking information on the velocity, position, and direction of wrist trajectory, but containing instead signals pertaining to spindle afferent discharge and tremor. Other evidence on cerebellar connectivity and the motor deficits after cerebellar ablation also suggest a control of properties of the motor apparatus rather than the programming of trajectory per se. By contrast, information on connectivity, motor deficits after ablation, and single unit recording during movement suggests that programming of trajectory might more properly be the role of the basal ganglia. A fourth set of experiments will test this hypothesis.